In spite of all we have learned about the genetic basis of colorectal cancer (CRC) over the past 25 years, a large proportion of patients with some form of familial CRC (FCC) cannot be characterized in a way that the family can be rationally counseled. This proposal is a dual-principal investigator renewal application for a highly productive program that has been ongoing for 18 years. Over the past 5 years, we have made progress in our understanding of how methylation-based silencing of the DNA mismatch repair (MMR) genes play a role in the genesis of CRC; our work has characterized early-onset CRCs that occur in the absence of FCC, and we found that genetic and epigenetic alterations in the tumor tissues of these patients are similar to that of non- Lynch syndrome FCC (or, FCC-type X). We characterized how alterations in expression of the MSH3 gene are involved in the genesis of sporadic CRCs. We have provided insights into the role of epigenetics in the genesis and clinical behavior of CRCs. The overarching goal of this program is to move toward a more complete personalization of the concepts used in the classification of familiality in CRC. We have four specific aims. Based upon our work on the somatic inactivation of MSH3 in CRCs, we propose to test the hypothesis that germline mutations in the MSH3 gene cause a previously unidentified disease - Lynch syndrome-MSH3 type, which would have a phenotype that would be systematically overlooked because of how we currently screen for Lynch syndrome. Second, we plan to develop an approach for finding balanced inversions in the Lynch syndrome-associated MMR genes to resolve the uncertainty in families where the anticipated germline mutations cannot be identified. Third, we propose that FCC-type X is associated with a unique tumor phenotype that will lead us to the germline basis of that disease. Finally, we propose to use whole genome sequencing techniques to test the hypothesis that the serrated polyposis syndrome (SPS) is a rare recessive disease caused by biallelic mutations in a gene involved in the methylator pathway of colorectal tumorigenesis. If successful, this work will fill several remaining gaps in the FCC problem, and is likely to have a substantial impact on patient care.